AutoML Leaderboard
| Best model | name | model_type | metric_type | metric_value | train_time | single_prediction_time |
|---|---|---|---|---|---|---|
| 1_DecisionTree | Decision Tree | logloss | 0.11699 | 57.1 | 0.1464 | |
| 2_DecisionTree | Decision Tree | logloss | 0.105097 | 50.46 | 0.149 | |
| 3_DecisionTree | Decision Tree | logloss | 0.105097 | 50.77 | 0.1449 | |
| 4_Default_Xgboost | Xgboost | logloss | 0.0789783 | 31.27 | 0.1474 | |
| 5_Default_RandomForest | Random Forest | logloss | 0.101761 | 34.37 | 0.8934 | |
| 6_Xgboost | Xgboost | logloss | 0.0793339 | 33.1 | 0.1455 | |
| 10_RandomForest | Random Forest | logloss | 0.105246 | 29.34 | 0.2903 | |
| 7_Xgboost | Xgboost | logloss | 0.0791386 | 29.91 | 0.1467 | |
| 11_RandomForest | Random Forest | logloss | 0.0854321 | 65.18 | 0.3418 | |
| 8_Xgboost | Xgboost | logloss | 0.0807353 | 42.92 | 0.1541 | |
| 12_RandomForest | Random Forest | logloss | 0.112054 | 28.35 | 0.2675 | |
| 9_Xgboost | Xgboost | logloss | 0.0839567 | 34.27 | 0.1571 | |
| 13_RandomForest | Random Forest | logloss | 0.0873966 | 51.14 | 0.3929 | |
| 4_Default_Xgboost_GoldenFeatures | Xgboost | logloss | 0.0793227 | 40.76 | 0.2174 | |
| 7_Xgboost_GoldenFeatures | Xgboost | logloss | 0.0797409 | 36.49 | 0.2137 | |
| 6_Xgboost_GoldenFeatures | Xgboost | logloss | 0.0797254 | 38.28 | 0.2196 | |
| the best | Ensemble | Ensemble | logloss | 0.0789624 | 10.64 | 0.309 |
AutoML Performance
AutoML Performance Boxplot
Features Importance
Spearman Correlation of Models
Summary of 13_RandomForest
Random Forest
- n_jobs: -1
- criterion: entropy
- max_features: 0.6
- min_samples_split: 50
- max_depth: 6
- eval_metric_name: logloss
- explain_level: 1
Validation
- validation_type: kfold
- k_folds: 5
- shuffle: True
- stratify: True
Optimized metric
logloss
Training time
49.2 seconds
Metric details
| score | threshold | |
|---|---|---|
| logloss | 0.0873966 | nan |
| auc | 0.973147 | nan |
| f1 | 0.798442 | 0.271141 |
| accuracy | 0.9709 | 0.271141 |
| precision | 1 | 0.995046 |
| recall | 1 | 0 |
| mcc | 0.797955 | 0.271141 |
Metric details with threshold from accuracy metric
| score | threshold | |
|---|---|---|
| logloss | 0.0873966 | nan |
| auc | 0.973147 | nan |
| f1 | 0.798442 | 0.271141 |
| accuracy | 0.9709 | 0.271141 |
| precision | 0.970737 | 0.271141 |
| recall | 0.678088 | 0.271141 |
| mcc | 0.797955 | 0.271141 |
Confusion matrix (at threshold=0.271141)
| Predicted as 0 | Predicted as 1 | |
|---|---|---|
| Labeled as 0 | 73061 | 139 |
| Labeled as 1 | 2189 | 4611 |
Learning curves
Permutation-based Importance
Confusion Matrix
Normalized Confusion Matrix
ROC Curve
Kolmogorov-Smirnov Statistic
Precision-Recall Curve
Calibration Curve
Cumulative Gains Curve
Lift Curve
Summary of 8_Xgboost
Extreme Gradient Boosting (Xgboost)
- n_jobs: -1
- objective: binary:logistic
- eta: 0.1
- max_depth: 7
- min_child_weight: 25
- subsample: 0.9
- colsample_bytree: 0.6
- eval_metric: logloss
- explain_level: 1
Validation
- validation_type: kfold
- k_folds: 5
- shuffle: True
- stratify: True
Optimized metric
logloss
Training time
41.1 seconds
Metric details
| score | threshold | |
|---|---|---|
| logloss | 0.0807353 | nan |
| auc | 0.97919 | nan |
| f1 | 0.806449 | 0.479264 |
| accuracy | 0.971938 | 0.479264 |
| precision | 1 | 0.972692 |
| recall | 1 | 0.00014167 |
| mcc | 0.805801 | 0.479264 |
Metric details with threshold from accuracy metric
| score | threshold | |
|---|---|---|
| logloss | 0.0807353 | nan |
| auc | 0.97919 | nan |
| f1 | 0.806449 | 0.479264 |
| accuracy | 0.971938 | 0.479264 |
| precision | 0.974578 | 0.479264 |
| recall | 0.687794 | 0.479264 |
| mcc | 0.805801 | 0.479264 |
Confusion matrix (at threshold=0.479264)
| Predicted as 0 | Predicted as 1 | |
|---|---|---|
| Labeled as 0 | 73078 | 122 |
| Labeled as 1 | 2123 | 4677 |
Learning curves
Permutation-based Importance
Confusion Matrix
Normalized Confusion Matrix
ROC Curve
Kolmogorov-Smirnov Statistic
Precision-Recall Curve
Calibration Curve
Cumulative Gains Curve
Lift Curve
Summary of 7_Xgboost
Extreme Gradient Boosting (Xgboost)
- n_jobs: -1
- objective: binary:logistic
- eta: 0.1
- max_depth: 8
- min_child_weight: 1
- subsample: 1.0
- colsample_bytree: 1.0
- eval_metric: logloss
- explain_level: 1
Validation
- validation_type: kfold
- k_folds: 5
- shuffle: True
- stratify: True
Optimized metric
logloss
Training time
28.1 seconds
Metric details
| score | threshold | |
|---|---|---|
| logloss | 0.0791386 | nan |
| auc | 0.979411 | nan |
| f1 | 0.805932 | 0.484127 |
| accuracy | 0.971862 | 0.484127 |
| precision | 1 | 0.99888 |
| recall | 1 | 2.58249e-05 |
| mcc | 0.805234 | 0.484127 |
Metric details with threshold from accuracy metric
| score | threshold | |
|---|---|---|
| logloss | 0.0791386 | nan |
| auc | 0.979411 | nan |
| f1 | 0.805932 | 0.484127 |
| accuracy | 0.971862 | 0.484127 |
| precision | 0.973953 | 0.484127 |
| recall | 0.687353 | 0.484127 |
| mcc | 0.805234 | 0.484127 |
Confusion matrix (at threshold=0.484127)
| Predicted as 0 | Predicted as 1 | |
|---|---|---|
| Labeled as 0 | 73075 | 125 |
| Labeled as 1 | 2126 | 4674 |
Learning curves
Permutation-based Importance
Confusion Matrix
Normalized Confusion Matrix
ROC Curve
Kolmogorov-Smirnov Statistic
Precision-Recall Curve
Calibration Curve
Cumulative Gains Curve
Lift Curve
Summary of 2_DecisionTree
Decision Tree
- n_jobs: -1
- criterion: gini
- max_depth: 4
- explain_level: 1
Validation
- validation_type: kfold
- k_folds: 5
- shuffle: True
- stratify: True
Optimized metric
logloss
Training time
48.8 seconds
Metric details
| score | threshold | |
|---|---|---|
| logloss | 0.105097 | nan |
| auc | 0.936293 | nan |
| f1 | 0.800811 | 0.112737 |
| accuracy | 0.971762 | 0.112737 |
| precision | 1 | 0.112737 |
| recall | 1 | 0 |
| mcc | 0.804862 | 0.112737 |
Metric details with threshold from accuracy metric
| score | threshold | |
|---|---|---|
| logloss | 0.105097 | nan |
| auc | 0.936293 | nan |
| f1 | 0.800811 | 0.112737 |
| accuracy | 0.971762 | 0.112737 |
| precision | 1 | 0.112737 |
| recall | 0.667794 | 0.112737 |
| mcc | 0.804862 | 0.112737 |
Confusion matrix (at threshold=0.112737)
| Predicted as 0 | Predicted as 1 | |
|---|---|---|
| Labeled as 0 | 73200 | 0 |
| Labeled as 1 | 2259 | 4541 |
Learning curves
Permutation-based Importance
Confusion Matrix
Normalized Confusion Matrix
ROC Curve
Kolmogorov-Smirnov Statistic
Precision-Recall Curve
Calibration Curve
Cumulative Gains Curve
Lift Curve
Summary of 4_Default_Xgboost_GoldenFeatures
Extreme Gradient Boosting (Xgboost)
- n_jobs: -1
- objective: binary:logistic
- eta: 0.075
- max_depth: 6
- min_child_weight: 1
- subsample: 1.0
- colsample_bytree: 1.0
- eval_metric: logloss
- explain_level: 1
Validation
- validation_type: kfold
- k_folds: 5
- shuffle: True
- stratify: True
Optimized metric
logloss
Training time
38.6 seconds
Metric details
| score | threshold | |
|---|---|---|
| logloss | 0.0793227 | nan |
| auc | 0.979189 | nan |
| f1 | 0.807828 | 0.485747 |
| accuracy | 0.972137 | 0.485747 |
| precision | 1 | 0.998885 |
| recall | 1 | 4.96688e-06 |
| mcc | 0.807311 | 0.485747 |
Metric details with threshold from accuracy metric
| score | threshold | |
|---|---|---|
| logloss | 0.0793227 | nan |
| auc | 0.979189 | nan |
| f1 | 0.807828 | 0.485747 |
| accuracy | 0.972137 | 0.485747 |
| precision | 0.976245 | 0.485747 |
| recall | 0.688971 | 0.485747 |
| mcc | 0.807311 | 0.485747 |
Confusion matrix (at threshold=0.485747)
| Predicted as 0 | Predicted as 1 | |
|---|---|---|
| Labeled as 0 | 73086 | 114 |
| Labeled as 1 | 2115 | 4685 |
Learning curves
Permutation-based Importance
Confusion Matrix
Normalized Confusion Matrix
ROC Curve
Kolmogorov-Smirnov Statistic
Precision-Recall Curve
Calibration Curve
Cumulative Gains Curve
Lift Curve
Summary of 5_Default_RandomForest
Random Forest
- n_jobs: -1
- criterion: gini
- max_features: 0.9
- min_samples_split: 30
- max_depth: 4
- eval_metric_name: logloss
- explain_level: 1
Validation
- validation_type: kfold
- k_folds: 5
- shuffle: True
- stratify: True
Optimized metric
logloss
Training time
32.6 seconds
Metric details
| score | threshold | |
|---|---|---|
| logloss | 0.101761 | nan |
| auc | 0.958443 | nan |
| f1 | 0.800811 | 0.113843 |
| accuracy | 0.971762 | 0.113843 |
| precision | 1 | 0.113843 |
| recall | 1 | 0 |
| mcc | 0.804862 | 0.113843 |
Metric details with threshold from accuracy metric
| score | threshold | |
|---|---|---|
| logloss | 0.101761 | nan |
| auc | 0.958443 | nan |
| f1 | 0.800811 | 0.113843 |
| accuracy | 0.971762 | 0.113843 |
| precision | 1 | 0.113843 |
| recall | 0.667794 | 0.113843 |
| mcc | 0.804862 | 0.113843 |
Confusion matrix (at threshold=0.113843)
| Predicted as 0 | Predicted as 1 | |
|---|---|---|
| Labeled as 0 | 73200 | 0 |
| Labeled as 1 | 2259 | 4541 |
Learning curves
Permutation-based Importance
Confusion Matrix
Normalized Confusion Matrix
ROC Curve
Kolmogorov-Smirnov Statistic
Precision-Recall Curve
Calibration Curve
Cumulative Gains Curve
Lift Curve
Summary of 7_Xgboost_GoldenFeatures
Extreme Gradient Boosting (Xgboost)
- n_jobs: -1
- objective: binary:logistic
- eta: 0.1
- max_depth: 8
- min_child_weight: 1
- subsample: 1.0
- colsample_bytree: 1.0
- eval_metric: logloss
- explain_level: 1
Validation
- validation_type: kfold
- k_folds: 5
- shuffle: True
- stratify: True
Optimized metric
logloss
Training time
34.6 seconds
Metric details
| score | threshold | |
|---|---|---|
| logloss | 0.0797409 | nan |
| auc | 0.978971 | nan |
| f1 | 0.806104 | 0.496844 |
| accuracy | 0.971888 | 0.496844 |
| precision | 1 | 0.998948 |
| recall | 1 | 1.72534e-05 |
| mcc | 0.805423 | 0.496844 |
Metric details with threshold from accuracy metric
| score | threshold | |
|---|---|---|
| logloss | 0.0797409 | nan |
| auc | 0.978971 | nan |
| f1 | 0.806104 | 0.496844 |
| accuracy | 0.971888 | 0.496844 |
| precision | 0.974161 | 0.496844 |
| recall | 0.6875 | 0.496844 |
| mcc | 0.805423 | 0.496844 |
Confusion matrix (at threshold=0.496844)
| Predicted as 0 | Predicted as 1 | |
|---|---|---|
| Labeled as 0 | 73076 | 124 |
| Labeled as 1 | 2125 | 4675 |
Learning curves
Permutation-based Importance
Confusion Matrix
Normalized Confusion Matrix
ROC Curve
Kolmogorov-Smirnov Statistic
Precision-Recall Curve
Calibration Curve
Cumulative Gains Curve
Lift Curve
Summary of 4_Default_Xgboost
Extreme Gradient Boosting (Xgboost)
- n_jobs: -1
- objective: binary:logistic
- eta: 0.075
- max_depth: 6
- min_child_weight: 1
- subsample: 1.0
- colsample_bytree: 1.0
- eval_metric: logloss
- explain_level: 1
Validation
- validation_type: kfold
- k_folds: 5
- shuffle: True
- stratify: True
Optimized metric
logloss
Training time
29.5 seconds
Metric details
| score | threshold | |
|---|---|---|
| logloss | 0.0789783 | nan |
| auc | 0.979491 | nan |
| f1 | 0.806966 | 0.471158 |
| accuracy | 0.972012 | 0.471158 |
| precision | 1 | 0.999139 |
| recall | 1 | 7.27766e-06 |
| mcc | 0.806367 | 0.471158 |
Metric details with threshold from accuracy metric
| score | threshold | |
|---|---|---|
| logloss | 0.0789783 | nan |
| auc | 0.979491 | nan |
| f1 | 0.806966 | 0.471158 |
| accuracy | 0.972012 | 0.471158 |
| precision | 0.975203 | 0.471158 |
| recall | 0.688235 | 0.471158 |
| mcc | 0.806367 | 0.471158 |
Confusion matrix (at threshold=0.471158)
| Predicted as 0 | Predicted as 1 | |
|---|---|---|
| Labeled as 0 | 73081 | 119 |
| Labeled as 1 | 2120 | 4680 |
Learning curves
Permutation-based Importance
Confusion Matrix
Normalized Confusion Matrix
ROC Curve
Kolmogorov-Smirnov Statistic
Precision-Recall Curve
Calibration Curve
Cumulative Gains Curve
Lift Curve
Summary of 12_RandomForest
Random Forest
- n_jobs: -1
- criterion: gini
- max_features: 0.8
- min_samples_split: 40
- max_depth: 3
- eval_metric_name: logloss
- explain_level: 1
Validation
- validation_type: kfold
- k_folds: 5
- shuffle: True
- stratify: True
Optimized metric
logloss
Training time
26.6 seconds
Metric details
| score | threshold | |
|---|---|---|
| logloss | 0.112054 | nan |
| auc | 0.93459 | nan |
| f1 | 0.800811 | 0.078801 |
| accuracy | 0.971762 | 0.078801 |
| precision | 1 | 0.078801 |
| recall | 1 | 0.00599173 |
| mcc | 0.804862 | 0.078801 |
Metric details with threshold from accuracy metric
| score | threshold | |
|---|---|---|
| logloss | 0.112054 | nan |
| auc | 0.93459 | nan |
| f1 | 0.800811 | 0.078801 |
| accuracy | 0.971762 | 0.078801 |
| precision | 1 | 0.078801 |
| recall | 0.667794 | 0.078801 |
| mcc | 0.804862 | 0.078801 |
Confusion matrix (at threshold=0.078801)
| Predicted as 0 | Predicted as 1 | |
|---|---|---|
| Labeled as 0 | 73200 | 0 |
| Labeled as 1 | 2259 | 4541 |
Learning curves
Permutation-based Importance
Confusion Matrix
Normalized Confusion Matrix
ROC Curve
Kolmogorov-Smirnov Statistic
Precision-Recall Curve
Calibration Curve
Cumulative Gains Curve
Lift Curve
Summary of 9_Xgboost
Extreme Gradient Boosting (Xgboost)
- n_jobs: -1
- objective: binary:logistic
- eta: 0.15
- max_depth: 8
- min_child_weight: 50
- subsample: 0.6
- colsample_bytree: 0.6
- eval_metric: logloss
- explain_level: 1
Validation
- validation_type: kfold
- k_folds: 5
- shuffle: True
- stratify: True
Optimized metric
logloss
Training time
32.3 seconds
Metric details
| score | threshold | |
|---|---|---|
| logloss | 0.0839567 | nan |
| auc | 0.977695 | nan |
| f1 | 0.805414 | 0.489164 |
| accuracy | 0.971788 | 0.489164 |
| precision | 1 | 0.933093 |
| recall | 1 | 0.0001576 |
| mcc | 0.804668 | 0.489164 |
Metric details with threshold from accuracy metric
| score | threshold | |
|---|---|---|
| logloss | 0.0839567 | nan |
| auc | 0.977695 | nan |
| f1 | 0.805414 | 0.489164 |
| accuracy | 0.971788 | 0.489164 |
| precision | 0.973328 | 0.489164 |
| recall | 0.686912 | 0.489164 |
| mcc | 0.804668 | 0.489164 |
Confusion matrix (at threshold=0.489164)
| Predicted as 0 | Predicted as 1 | |
|---|---|---|
| Labeled as 0 | 73072 | 128 |
| Labeled as 1 | 2129 | 4671 |
Learning curves
Permutation-based Importance
Confusion Matrix
Normalized Confusion Matrix
ROC Curve
Kolmogorov-Smirnov Statistic
Precision-Recall Curve
Calibration Curve
Cumulative Gains Curve
Lift Curve
Summary of 11_RandomForest
Random Forest
- n_jobs: -1
- criterion: gini
- max_features: 0.7
- min_samples_split: 30
- max_depth: 7
- eval_metric_name: logloss
- explain_level: 1
Validation
- validation_type: kfold
- k_folds: 5
- shuffle: True
- stratify: True
Optimized metric
logloss
Training time
63.4 seconds
Metric details
| score | threshold | |
|---|---|---|
| logloss | 0.0854321 | nan |
| auc | 0.974706 | nan |
| f1 | 0.805829 | 0.304323 |
| accuracy | 0.97185 | 0.304323 |
| precision | 0.973947 | 0.304323 |
| recall | 1 | 0 |
| mcc | 0.80514 | 0.304323 |
Metric details with threshold from accuracy metric
| score | threshold | |
|---|---|---|
| logloss | 0.0854321 | nan |
| auc | 0.974706 | nan |
| f1 | 0.805829 | 0.304323 |
| accuracy | 0.97185 | 0.304323 |
| precision | 0.973947 | 0.304323 |
| recall | 0.687206 | 0.304323 |
| mcc | 0.80514 | 0.304323 |
Confusion matrix (at threshold=0.304323)
| Predicted as 0 | Predicted as 1 | |
|---|---|---|
| Labeled as 0 | 73075 | 125 |
| Labeled as 1 | 2127 | 4673 |
Learning curves
Permutation-based Importance
Confusion Matrix
Normalized Confusion Matrix
ROC Curve
Kolmogorov-Smirnov Statistic
Precision-Recall Curve
Calibration Curve
Cumulative Gains Curve
Lift Curve
Summary of 1_DecisionTree
Decision Tree
- n_jobs: -1
- criterion: gini
- max_depth: 3
- explain_level: 1
Validation
- validation_type: kfold
- k_folds: 5
- shuffle: True
- stratify: True
Optimized metric
logloss
Training time
55.5 seconds
Metric details
| score | threshold | |
|---|---|---|
| logloss | 0.11699 | nan |
| auc | 0.902112 | nan |
| f1 | 0.800811 | 0.0694736 |
| accuracy | 0.971762 | 0.0694736 |
| precision | 1 | 0.0694736 |
| recall | 1 | 0.00991861 |
| mcc | 0.804862 | 0.0694736 |
Metric details with threshold from accuracy metric
| score | threshold | |
|---|---|---|
| logloss | 0.11699 | nan |
| auc | 0.902112 | nan |
| f1 | 0.800811 | 0.0694736 |
| accuracy | 0.971762 | 0.0694736 |
| precision | 1 | 0.0694736 |
| recall | 0.667794 | 0.0694736 |
| mcc | 0.804862 | 0.0694736 |
Confusion matrix (at threshold=0.069474)
| Predicted as 0 | Predicted as 1 | |
|---|---|---|
| Labeled as 0 | 73200 | 0 |
| Labeled as 1 | 2259 | 4541 |
Learning curves
Permutation-based Importance
Confusion Matrix
Normalized Confusion Matrix
ROC Curve
Kolmogorov-Smirnov Statistic
Precision-Recall Curve
Calibration Curve
Cumulative Gains Curve
Lift Curve
Summary of 6_Xgboost
Extreme Gradient Boosting (Xgboost)
- n_jobs: -1
- objective: binary:logistic
- eta: 0.075
- max_depth: 8
- min_child_weight: 5
- subsample: 1.0
- colsample_bytree: 1.0
- eval_metric: logloss
- explain_level: 1
Validation
- validation_type: kfold
- k_folds: 5
- shuffle: True
- stratify: True
Optimized metric
logloss
Training time
31.3 seconds
Metric details
| score | threshold | |
|---|---|---|
| logloss | 0.0793339 | nan |
| auc | 0.97933 | nan |
| f1 | 0.806966 | 0.476171 |
| accuracy | 0.972012 | 0.476171 |
| precision | 1 | 0.996978 |
| recall | 1 | 4.94245e-05 |
| mcc | 0.806367 | 0.476171 |
Metric details with threshold from accuracy metric
| score | threshold | |
|---|---|---|
| logloss | 0.0793339 | nan |
| auc | 0.97933 | nan |
| f1 | 0.806966 | 0.476171 |
| accuracy | 0.972012 | 0.476171 |
| precision | 0.975203 | 0.476171 |
| recall | 0.688235 | 0.476171 |
| mcc | 0.806367 | 0.476171 |
Confusion matrix (at threshold=0.476171)
| Predicted as 0 | Predicted as 1 | |
|---|---|---|
| Labeled as 0 | 73081 | 119 |
| Labeled as 1 | 2120 | 4680 |
Learning curves
Permutation-based Importance
Confusion Matrix
Normalized Confusion Matrix
ROC Curve
Kolmogorov-Smirnov Statistic
Precision-Recall Curve
Calibration Curve
Cumulative Gains Curve
Lift Curve
Summary of 3_DecisionTree
Decision Tree
- n_jobs: -1
- criterion: gini
- max_depth: 4
- explain_level: 1
Validation
- validation_type: kfold
- k_folds: 5
- shuffle: True
- stratify: True
Optimized metric
logloss
Training time
48.6 seconds
Metric details
| score | threshold | |
|---|---|---|
| logloss | 0.105097 | nan |
| auc | 0.936293 | nan |
| f1 | 0.800811 | 0.112737 |
| accuracy | 0.971762 | 0.112737 |
| precision | 1 | 0.112737 |
| recall | 1 | 0 |
| mcc | 0.804862 | 0.112737 |
Metric details with threshold from accuracy metric
| score | threshold | |
|---|---|---|
| logloss | 0.105097 | nan |
| auc | 0.936293 | nan |
| f1 | 0.800811 | 0.112737 |
| accuracy | 0.971762 | 0.112737 |
| precision | 1 | 0.112737 |
| recall | 0.667794 | 0.112737 |
| mcc | 0.804862 | 0.112737 |
Confusion matrix (at threshold=0.112737)
| Predicted as 0 | Predicted as 1 | |
|---|---|---|
| Labeled as 0 | 73200 | 0 |
| Labeled as 1 | 2259 | 4541 |
Learning curves
Permutation-based Importance
Confusion Matrix
Normalized Confusion Matrix
ROC Curve
Kolmogorov-Smirnov Statistic
Precision-Recall Curve
Calibration Curve
Cumulative Gains Curve
Lift Curve
Summary of Ensemble
Ensemble structure
| Model | Weight |
|---|---|
| 4_Default_Xgboost | 10 |
| 7_Xgboost | 3 |
Metric details
| score | threshold | |
|---|---|---|
| logloss | 0.0789624 | nan |
| auc | 0.979507 | nan |
| f1 | 0.806966 | 0.472006 |
| accuracy | 0.972012 | 0.472006 |
| precision | 1 | 0.99912 |
| recall | 1 | 1.23171e-05 |
| mcc | 0.806367 | 0.472006 |
Metric details with threshold from accuracy metric
| score | threshold | |
|---|---|---|
| logloss | 0.0789624 | nan |
| auc | 0.979507 | nan |
| f1 | 0.806966 | 0.472006 |
| accuracy | 0.972012 | 0.472006 |
| precision | 0.975203 | 0.472006 |
| recall | 0.688235 | 0.472006 |
| mcc | 0.806367 | 0.472006 |
Confusion matrix (at threshold=0.472006)
| Predicted as 0 | Predicted as 1 | |
|---|---|---|
| Labeled as 0 | 73081 | 119 |
| Labeled as 1 | 2120 | 4680 |
Learning curves
Confusion Matrix
Normalized Confusion Matrix
ROC Curve
Kolmogorov-Smirnov Statistic
Precision-Recall Curve
Calibration Curve
Cumulative Gains Curve
Lift Curve
Summary of 6_Xgboost_GoldenFeatures
Extreme Gradient Boosting (Xgboost)
- n_jobs: -1
- objective: binary:logistic
- eta: 0.075
- max_depth: 8
- min_child_weight: 5
- subsample: 1.0
- colsample_bytree: 1.0
- eval_metric: logloss
- explain_level: 1
Validation
- validation_type: kfold
- k_folds: 5
- shuffle: True
- stratify: True
Optimized metric
logloss
Training time
36.4 seconds
Metric details
| score | threshold | |
|---|---|---|
| logloss | 0.0797254 | nan |
| auc | 0.979014 | nan |
| f1 | 0.807828 | 0.494189 |
| accuracy | 0.972137 | 0.494189 |
| precision | 1 | 0.99741 |
| recall | 1 | 9.51015e-05 |
| mcc | 0.807311 | 0.494189 |
Metric details with threshold from accuracy metric
| score | threshold | |
|---|---|---|
| logloss | 0.0797254 | nan |
| auc | 0.979014 | nan |
| f1 | 0.807828 | 0.494189 |
| accuracy | 0.972137 | 0.494189 |
| precision | 0.976245 | 0.494189 |
| recall | 0.688971 | 0.494189 |
| mcc | 0.807311 | 0.494189 |
Confusion matrix (at threshold=0.494189)
| Predicted as 0 | Predicted as 1 | |
|---|---|---|
| Labeled as 0 | 73086 | 114 |
| Labeled as 1 | 2115 | 4685 |
Learning curves
Permutation-based Importance
Confusion Matrix
Normalized Confusion Matrix
ROC Curve
Kolmogorov-Smirnov Statistic
Precision-Recall Curve
Calibration Curve
Cumulative Gains Curve
Lift Curve
Summary of 10_RandomForest
Random Forest
- n_jobs: -1
- criterion: gini
- max_features: 0.5
- min_samples_split: 20
- max_depth: 4
- eval_metric_name: logloss
- explain_level: 1
Validation
- validation_type: kfold
- k_folds: 5
- shuffle: True
- stratify: True
Optimized metric
logloss
Training time
26.8 seconds
Metric details
| score | threshold | |
|---|---|---|
| logloss | 0.105246 | nan |
| auc | 0.955361 | nan |
| f1 | 0.802248 | 0.132446 |
| accuracy | 0.971413 | 0.132446 |
| precision | 1 | 0.890705 |
| recall | 1 | 0.00345914 |
| mcc | 0.801838 | 0.132446 |
Metric details with threshold from accuracy metric
| score | threshold | |
|---|---|---|
| logloss | 0.105246 | nan |
| auc | 0.955361 | nan |
| f1 | 0.802248 | 0.132446 |
| accuracy | 0.971413 | 0.132446 |
| precision | 0.973557 | 0.132446 |
| recall | 0.682206 | 0.132446 |
| mcc | 0.801838 | 0.132446 |
Confusion matrix (at threshold=0.132446)
| Predicted as 0 | Predicted as 1 | |
|---|---|---|
| Labeled as 0 | 73074 | 126 |
| Labeled as 1 | 2161 | 4639 |
Learning curves
Permutation-based Importance
Confusion Matrix
Normalized Confusion Matrix
ROC Curve
Kolmogorov-Smirnov Statistic
Precision-Recall Curve
Calibration Curve
Cumulative Gains Curve
Lift Curve
